Anatomy of an AI Deal

In today’s market, time-to-value for technology projects is critical. The abrupt halt in demand and sharp drop in oil prices has accelerated Tech adoption across the sector. While some Majors may still look for a comprehensive AI platform to enable their digital transformation program in-house, many of our Energy clients are turning to off-the-shelf AI Applications to fast-track operational and financial results delivery.

What are our clients solving with AI?

The type of problems addressable vary but can often be directly tackled by an already developed AI Application. As an example, a National Oil Company (NOC) recently expressed interest in our Inventory Optimization AI Application to better optimize spares in its upstream operations. Similarly, a global chemicals and refining company asked us to deploy our Reliability AI Application to optimize their major production lines.

However, other challenges are not directly addressable by any existing AI application and do require specific development work. Another NOC recently asked us to respond to a project covering stuck pipe prevention, a common drilling issue in the industry accounting for 25% of NPT globally. For years, this company tried to address this issue with various digital solutions leveraging essentially physics-based models, with results below expectations. No off-the-shelf AI Application exists yet exist to tackle this very specific use case and its development will only be possible by combining our cutting-edge AI/ML tech with our deep oilfield expertise, as well through active collaboration with our customers practitioners.

A dynamic dialogue with our key partners and customers over the past year has enabled us to sketch a fairly standard journey that underpins the deployment of AI in Organisations, depicted hereafter:

1- Start the journey with a production prototype trial, demonstrate value and prove AI works

Our client's AI journey often starts with a trial project, allowing to sample the benefits of integrating AI into business operations. This begins by identifying a strong opportunity for digital technology (or "use case”) to meet a specific business objective as well as evaluating the economic value at stake. For example, a refinery would be interested in reducing process upsets in OpEx. We then take historical operating data and identify anomalies, often requiring a combination of machine and human interaction Each time an anomaly is found, our team validates with the customer what occurred to cause that anomaly and creates a tag, allowing the machine learning algorithms to learn. Trials can take between 2 to 4 months and move into full production (deployment at scale and at pace across the organization) once the benefits are demonstrated.

2-?Select AI Applications and set-up a Software-as-a-Service (SaaS) contract

Once the trial is successfully completed, our customers often recognize more opportunities for digital improvements and opt to purchase our different AI Applications as a service. These types of contracts include data integration, application customization, testing and training before final deployment, require roughly 3 to 6 months per application and are referred to as Software as a Service (SaaS) contracts.

3- Implement AI through a Digital Center of Excellence (CoE)

If most of an organization’s AI needs can't be addressed by existing AI applications, our customers often opt to set up an in-house AI CoE to fully leverage our AI platform. Time to value for these large platforms and new application development projects is longer - typically between 1 to 3 years - and requires key stakeholders within the customer’s organization to usher the process of integrating AI to their operations. These stakeholders support the effort with extensive change management to enable the adoption and adjustment of underlying processes and management systems.?

In each of the above steps, our BakerHughesC3.ai people are embracing best practices to deliver AI solutions, ensure success in the face of limited resources, manage a rapidly changing technology landscape and above all deliver significant economic value in line with our customer's expectations.

Capitalizing on the power of AI, they are improving efficiencies across operations....